All cathodes in the Magnavox amp being built sit just a few volts above ground, so there is marginal benefit of up-biasing the filaments. The benefit gained by up-biasing the filaments is the "reverse bias" condition it creates between heater and cathode, so that in a case of a tube with marginal heater insulation, current cannot flow from the filament to the cathode and introduce hum. That appears to have been the thinking in the original design of this amp.
If memory serves, the power amp sections of the Fisher models you mention have the inverter triode's cathode sitting at something like 100V DC above ground potential. In this scenario, in order to stay away from the max heater-cathode limits, and to alleviate a bit the stress on the heater insulation, it's wise to up-bias the filaments by "some appropriate" amount. There are two limits that must be managed when biasing filaments: the "heater positive with respect to cathode" limit and the "heater negative with respect to cathode" limit. You don't want to up-bias too far or you will run into the heater positive with respect to cathode limit for those cathodes that run just a few volts above ground. So what is usually done is to "split the difference" somehow. What I usually do is divide the highest voltage cathode by 4 (that turns out to be about 25V in situations like the Fisher models mentioned) and bias up all filaments that run on that filament secondary winding by that amount. Thus the tubes that run their cathodes a few volts above ground run at approximately +25V heater positive respect to cathode, while the phase inverter triode runs at approximately -75V heater negative respect to cathode. For a 12AX7, those voltages are well within range of both of the heater/cathode limits.
Before you bias the filaments though, it's always good to validate the heater-cathode ratings (both directions) to ensure you are well within the limits.